Robot Visions (Robot 0.5)

Essays The Friends We Make

The term "robot" dates back only sixty years. It was invented by the Czech playwright, Karel Capek, in his play, R. U. R., and is a Czech word meaning worker.

The idea, however, is far older. It is as old as man's longing for a servant as smart as a human being, but far stronger, and incapable of growing weary, bored, or dissatisfied. In the Greek myths, the god of the forge, Hephaistos, had two golden girls-as bright and alive as flesh-and-blood girls-to help him. And the island of Crete was guarded, in the myths, by a bronze giant named Talos, who circled its shores perpetually and tirelessly, watching for intruders.

Are robots possible, though? And if they are, are they desirable?

Mechanical devices with gears and springs and ratchets could certainly make manlike devices perform manlike actions, but the essence of a successful robot is to have it think-and think well enough to perform useful functions without being continually supervised.

But thinking takes a brain. The human being is made up of microscopic neurons, each of which has an extraordinarily complex substructure. There are 10 billion neurons in the brain and 90 billion supporting cells, all hooked together in a very intricate pattern. How can anything like that be duplicated by some man-made device in a robot?

It wasn't until the invention of the electronic computer thirty-five years ago that such a thing became conceivable. Since its birth, the electronic computer has grown ever more compact, and each year it becomes possible to pack more and more information into less and less volume.

In a few decades, might not enough versatility to direct a robot be packed into a volume the size of the human brain? Such a computer would not have to be as advanced as the human brain, but only advanced enough to guide the actions of a robot designed, let us say, to vacuum rugs, to run a hydraulic press, to survey the lunar surface.

A robot would, of course, have to include a self-contained energy source; we couldn't expect it to be forever plugged into a wall socket. This, however, can be handled. A battery that needs periodic charging is not so different from a living body that needs periodic feeding.

But why bother with a humanoid shape? Would it not be more sensible to devise a specialized machine to perform a particular task without asking it to take on all the inefficiencies involved in arms, legs, and torso? Suppose you design a robot that can hold a finger in a furnace to test its temperature and turn the heating unit on and off to maintain that temperature nearly constant. Surely a simple thermostat made of a bimetallic strip will do the job as well.

Consider, though, that over the thousands of years of man's civilization, we have built a technology geared to the human shape. Products for humans' use are designed in size and form to accommodate the human body-how it bends and how long, wide, and heavy the various bending parts are. Machines are designed to fit the human reach and the width and position of human fingers.

We have only to consider the problems of human beings who happen to be a little taller or shorter than the norm-or even just left-handed-to see how important it is to have a good fit into our technology.

If we want a directing device then, one that can make use of human tools and machines, and that can fit into the technology, we would find it useful to make that device in the human shape, with all the bends and turns of which the human body is capable. Nor would we want it to be too heavy or too abnormally proportioned. Average in all respects would be best.

Then too, we relate to all nonhuman things by finding, or inventing, something human about them. We attribute human characteristics to our pets, and even to our automobiles. We personify nature and all the products of nature and, in earlier times, made human-shaped gods and goddesses out of them.

Surely, if we are to take on thinking partners-or, at the least, thinking servants-in the form of machines, we will be more comfortable with them, and we will relate to them more easily, if they are shaped like humans.

It will be easier to be friends with human-shaped robots than with specialized machines of unrecognizable shape. And I sometimes think that, in the desperate straits of humanity today, we would be grateful to have nonhuman friends, even if they are only friends we build ourselves.

***